摘要
结合钢板制造过程中常见的表面缺陷,提出了一种基于分水岭变换的图像分割方法。采用数学形态学的重构算子,对原图像的梯度进行同态修正,将修正后的图像作为分水岭算法的输入。通过仿真,这种方法能够准确提取出钢板表面的缺陷位置,并有效控制了过分割现象的产生。
This paper proposes an image segmentation method based on watershed transformation,aiming at extracting common surface defects on steel sheets.This paper used reconstruction operator of mathematical morphology to correct gradients of the original image,and introduces the image after correction as the input of watershed algorithm.The simulations proved that this method can accurately extract the locations of surface defects on steel sheets.
出处
《工业控制计算机》
2016年第5期125-126,共2页
Industrial Control Computer
基金
中央高校基本科研业务费专项资金资助(163111005)
关键词
图像分割
钢板表面缺陷
分水岭算法
形态重构
过分割
image segmentation
surface defects on steel sheets
watershed algorithm
reconstruction of form
over-segmentation problem
